Abstract
It is commonly thought that the AP is determined by high frequency units in the most basal part of the cochlea, and that the frequency specificity of this response is rather poor. We have evaluated these relations for the 7 msec far field activity by means of masking experiments and recordings from the Vertex. Stimuli were tone bursts at 1, 2 and 4 kHz with rise-fall times from 1 to 3 msec. Peak duration 0.5 msec. Maskers were noisebands with center frequencies spaced at narrow intervals from 0.5 to 6 kHz. In all situations where the frequency for the masker and the stimulus were close to each other there were pronounced and characteristic changes in the response. The results clearly demonstrate that the 7 msec response reflects activity in the cochlea that is characteristic for each stimulus frequency. We were unable to detect any particular influence from the most basal part of the cochlea. There is thus a distinct discrepancy between the frequency selectivity for the AP and for the brainstem responses. Because of the particular mechanisms in the cochlea the compound potentials that are recorded from the afferent acoustic system represent a convolution over time of unit responses that are generated successively along the cochlea partition. Events in the time domain should be the same at higher and at lower recording stations, but the duration of the unit response may be different. If we presume that unit responses in the higher centres last longer they would not require the same degree of synchronization for the generation of a compound potential. It is the speed of the travelling wave that determines the degree of synchronization and it decreases logarithmically towards the apex of the cochlea. Thus it may be understandable that brainstem nuclei and the acoustic nerve differ with regard to the generation of compound potentials.